Worried about a Trump loss? Don't.
Post Reply
 
Thread Rating:
  • 0 Votes - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
21-08-2016, 10:21 AM
RE: Worried about a Trump loss? Don't.
(21-08-2016 09:31 AM)GirlyMan Wrote:  And let's not forget his reach out to black voters. "You're living in your poverty, your schools are no good, you have no jobs, 58 percent of your youth is unemployed — what the hell do you have to lose?"




Facepalm What a farce he is.
Find all posts by this user
Like Post Quote this message in a reply
[+] 2 users Like Fireball's post
21-08-2016, 10:30 AM
RE: Worried about a Trump loss? Don't.
(21-08-2016 10:16 AM)GirlyMan Wrote:  
(21-08-2016 10:07 AM)Vosur Wrote:  [Image: zAFxOg.png]

Source: http://graphics.latimes.com/usc-presiden...dashboard/

Consider

Told you there was nothing to worry about. He's got a lock on the negro vote too.

Are you saying that the majority of blacks would rather have an orange man as president instead of a white woman?

See here they are the bruises some were self-inflicted and some showed up along the way. - JF

We're all mad here. The Cheshire Cat
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes Anjele's post
21-08-2016, 10:31 AM
RE: Worried about a Trump loss? Don't.
(21-08-2016 10:21 AM)Fireball Wrote:  
(21-08-2016 09:31 AM)GirlyMan Wrote:  And let's not forget his reach out to black voters. "You're living in your poverty, your schools are no good, you have no jobs, 58 percent of your youth is unemployed — what the hell do you have to lose?"
Facepalm What a farce he is.

A charade even.




#sigh
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes GirlyMan's post
21-08-2016, 08:35 PM (This post was last modified: 21-08-2016 08:38 PM by Full Circle.)
RE: Worried about a Trump loss? Don't.
(21-08-2016 10:07 AM)Vosur Wrote:  
(21-08-2016 09:31 AM)GirlyMan Wrote:  And let's not forget his reach out to black voters. "You're living in your poverty, your schools are no good, you have no jobs, 58 percent of your youth is unemployed — what the hell do you have to lose?"




[Image: zAFxOg.png]

Source: http://graphics.latimes.com/usc-presiden...dashboard/

Consider

The issue I take with the methodology for the results by this organization (and many others to be fair) is the sample size they use, in this case 3,000 individuals polled 400 at a time.

I think it is absurd to think that 400 individuals can fairly represent the entire voting population that has averaged about 130 million voters in the last two elections! The Law of Large Numbers is being ignored. In my estimation any poll that doesn’t even use 1% of the probable electorate is meaningless, and even 1% (1.3 million) might be too small considering a margin of error of 3% (they don’t say what that margin is that I could find).

Smaller sample sizes, just as smaller populations, are more variable than large ones.

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
21-08-2016, 08:43 PM
RE: Worried about a Trump loss? Don't.
(21-08-2016 08:35 PM)Full Circle Wrote:  
(21-08-2016 10:07 AM)Vosur Wrote:  [Image: zAFxOg.png]

Source: http://graphics.latimes.com/usc-presiden...dashboard/

Consider

The issue I take with the methodology for the results by this organization (and many others to be fair) is the sample size they use, in this case 3,000 individuals polled 400 at a time.

I think it is absurd to think that 400 individuals can fairly represent the entire voting population that has averaged about 130 million voters in the last two elections! The Law of Large Numbers is being ignored. In my estimation any poll that doesn’t even use 1% of the probable electorate is meaningless, and even 1% (1.3 million) might be too small considering a margin of error of 3% (they don’t say what that margin is that I could find).

Smaller sample sizes, just as smaller populations, are more variable than large ones.
It makes intuitive sense that a sample size that small couldn't possibly be representative of the whole, but that's exactly what the history of political polling shows us. If we require 1% of the population as a sample size for a poll to be meaningful, then every political poll that has ever been conducted is meaningless.

[Image: 7oDSbD4.gif]
Find all posts by this user
Like Post Quote this message in a reply
21-08-2016, 09:07 PM
RE: Worried about a Trump loss? Don't.
(21-08-2016 08:43 PM)Vosur Wrote:  It makes intuitive sense that a sample size that small couldn't possibly be representative of the whole, but that's exactly what the history of political polling shows us. If we require 1% of the population as a sample size for a poll to be meaningful, then every political poll that has ever been conducted is meaningless.

I would agree with the latter. This is why the latest 27 polls by 27 different organizations vary so greatly, none use a statistically significant sample size. Some are bound to be somewhat right, after all there can only be two outcomes, and because of this people will continue to pay small sample polls undue attention.

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
21-08-2016, 10:12 PM
RE: Worried about a Trump loss? Don't.
(21-08-2016 08:35 PM)Full Circle Wrote:  ... The issue I take with the methodology for the results by this organization ... is the sample size they use ...

The reason smaller than what seems reasonable sample sizes work is that the degree of meaningful variance between individuals drops off quickly, such that the differences between 500 people drawn from 300,000,000 and another 500 drawn from the same pool won't be significantly different from the differences between two groups of 5000 drawn from the 300,000,000 pool. You don't need the 300,000,000th person to account for all meaningful differences - you reach all the statistically meaningful variance well before then. And you can't do a "Trump" and go with "your gut" in ascertaining what constitutes an adequate sample size, you need scientifically conducted statistical studies to find it - and you can believe that the credible polling houses, and there are several (see Nate Silver for which they are) know EXACTLY what sample sizes produce meaningful data.
Find all posts by this user
Like Post Quote this message in a reply
21-08-2016, 11:02 PM (This post was last modified: 21-08-2016 11:50 PM by Full Circle.)
RE: Worried about a Trump loss? Don't.
(21-08-2016 10:12 PM)Airportkid Wrote:  
(21-08-2016 08:35 PM)Full Circle Wrote:  ... The issue I take with the methodology for the results by this organization ... is the sample size they use ...

The reason smaller than what seems reasonable sample sizes work is that the degree of meaningful variance between individuals drops off quickly, such that the differences between 500 people drawn from 300,000,000 and another 500 drawn from the same pool won't be significantly different from the differences between two groups of 5000 drawn from the 300,000,000 pool. You don't need the 300,000,000th person to account for all meaningful differences - you reach all the statistically meaningful variance well before then. And you can't do a "Trump" and go with "your gut" in ascertaining what constitutes an adequate sample size, you need scientifically conducted statistical studies to find it - and you can believe that the credible polling houses, and there are several (see Nate Silver for which they are) know EXACTLY what sample sizes produce meaningful data.

As the sample size grows the variability drops because it dilutes what has already happened, in this case the opinion of 400 people, until it is proportionally negligible.

Looked at another way, it is more likely to flip a coin 10 times and have it come up heads 9 or even 10 times than it is to flip it 1,000 times and have it come up 900-1000 times heads. That is the Law of Large Numbers.

400 people simply isn’t a statistically meaningful number to draw conclusions on how 130,000,000 people will vote.

PS Nate Silver and his 538 methods are very complex and at its core it is a “poll of polls” with all sorts of other tinkering. If you look here you’ll see which polls he uses most and I have no idea of their methodologies. My argument was strictly regarding the poll Vosur linked and its particular methodology.
http://projects.fivethirtyeight.com/pollster-ratings/

“I am quite sure now that often, very often, in matters concerning religion and politics a man’s reasoning powers are not above the monkey’s.”~Mark Twain
“Ocean: A body of water occupying about two-thirds of a world made for man - who has no gills.”~ Ambrose Bierce
Find all posts by this user
Like Post Quote this message in a reply
21-08-2016, 11:18 PM
RE: Worried about a Trump loss? Don't.
Quote:Looked at another way, it is more likely to flip a coin 10 times and have it come up heads 9 or even 10 times than it is to flip it 1,000 times and have it come up 900-1000 times heads. That is the Law of Large Numbers.
Thanks, Full Circle.

I try to learn something every day.

Tomorrow is precious, don't ruin it by fouling up today.
Find all posts by this user
Like Post Quote this message in a reply
22-08-2016, 09:43 AM
RE: Worried about a Trump loss? Don't.
(21-08-2016 08:35 PM)Full Circle Wrote:  The issue I take with the methodology for the results by this organization (and many others to be fair) is the sample size they use, in this case 3,000 individuals polled 400 at a time.

I think it is absurd to think that 400 individuals can fairly represent the entire voting population that has averaged about 130 million voters in the last two elections! The Law of Large Numbers is being ignored. In my estimation any poll that doesn’t even use 1% of the probable electorate is meaningless, and even 1% (1.3 million) might be too small considering a margin of error of 3% (they don’t say what that margin is that I could find).

Smaller sample sizes, just as smaller populations, are more variable than large ones.

This is how they do it. It's used all the time in clinical trials as well typically with much smaller sample sizes.

#sigh
Find all posts by this user
Like Post Quote this message in a reply
[+] 1 user Likes GirlyMan's post
Post Reply
Forum Jump: